Bisphosphonates Induce Apoptosis of Stromal Tumor Cells in Giant ...

2 downloads 0 Views 312KB Size Report
Giant cell tumour of bone (GCT) is an aggressive primary neoplasm that results in the production of osteolytic lesions. Stromal cells, which form the main ...
Calcif Tissue Int (2004) 75:71–77 DOI: 10.1007/s00223-004-0120-2

Bisphosphonates Induce Apoptosis of Stromal Tumor Cells in Giant Cell Tumor of Bone Y. Y. Cheng,1 L. Huang,1 K. M. Lee,1 J. K. Xu,2 M. H. Zheng,2 S. M. Kumta1 1 2

Department of Orthopaedics and Traumatology, Chinese University of Hong Kong, Shatin, N.T., Hong Kong SAR Department of Orthopaedic Surgery, University of Western Australia, QE II Medical Centre, Nedlands, Western Australia

Received: 23 May 2003 / Accepted: 29 December 2003 / Online publication: 25 March 2004

Abstract. Giant cell tumour of bone (GCT) is an aggressive primary neoplasm that results in the production of osteolytic lesions. Stromal cells, which form the main neoplastic component of this tumor, regulate the formation of osleoclast-like giant cells that are ultimately responsible for bone destruction. Bisphosphonates prevent bone resorption by inhibiting osteoclast activity and promoting osteoclast apoptosis, and they have been known to induce apoptosis of primary neoplastic cells such as those in breast and prostate cancers. We hypothesized that in bisphosphonates may induce apoptosis not only in osteoclast-like giant cells but also in neoplastic stromal cells of GCT both in vitro and in vivo. Twelve patients with GCT were treated with weekly injections of pamidronate for a period of 6 weeks prior to surgery. GCT specimens were collected at the time of biopsy and during definitive surgery. TUNEL assay was used to evaluate apoptotic DNA fragmentation in cells. In addition, twelve GCT primary cultures from these patients were treated with zoledronate, pamidronate, or alendronate for 48 hours at different doses (3, 30, or 150 lM) and subjected to apoptosis assay by flow cytometry following fluorescent Annexin-V labeling. The results showed that pamidronate significantly induced apoptosis in both osteoclast-like giant cells and stromal tumor cells, in vivo. All three bisphosphonates caused substantial apoptosis of stromal tumor cells in cultures. Zoledronate was the most potent reagent, resulting in an average cell death of 27.41% at 150 lM, followed by pamidronate (22.23%) and alendronate (15.3%). Our observations suggest that these drugs may be considered as potential adjuvants in the treatment of GCT. Key words: Giant cell tumor of bone (GCT) — Stromal cell — Bisphosphonates — Apoptosis

Giant cell tumor of bone (GCT) is a benign tumor that nevertheless exhibits considerable local aggressiveness and recurrence tendencies. It typically affects the Both authors (Y.Y. Cheng and L. Huang) contributed equally to this work Correspondence to: S. M. Kumta; E-mail: kumta@cuhk. edu.hk

epiphyseal ends of long bones such as the distal femur, the proximal tibia, and the distal radius, prompting the formation of a local osteolytic lesion [1]. Histologically GCT consists of three major cell types: a high number of osteoclast-like giant cells that destroy the affected bone, monocytic round cells, and spindle-shaped fibroblastlike stromal tumor cells [2, 3]. Cell culture experiments have established that stromal cells are the only proliferating components of GCT; monocytes and the osteoclast-like giant cells have been known to disappear after a few culture passages [4]. In this pathological environment, the neoplastic nature of the stromal component drives the hematopoietic precursors to undergo fusion and form osteoclasts, which eventually cause aggressive bone resorption and skeletal destruction [2, 5]. A high recurrence rate of 18% to 50% has been generally reported for GCT [6–8]. This tumor is known to metastasize and a potential for occasional malignant transformation has also been demonstrated [9]. Conventional chemotherapy has rarely been used for the treatment of GCT, primarily because of unacceptable toxicity for an otherwise nonfatal tumor. It is for these reasons that the surgical treatment of GCT poses a significant challenge, and an effort has been made in recent years to look for a less toxic and more successful alternative treatment modality or adjuvant therapy. Bisphosphonates are analogues of endogenous pyrophosphonate. In vivo, bisphosphonates bind strongly to hydroxyapatite on the bone surface and are thus preferentially delivered to sites of increased bone formation or resorption [10]. They are potent inhibitors of osteoclast-mediated bone resorption, a direct result of their ability to retard osteoclast generation and maturation while decreasing activity [11–13]. Bisphosphonates accomplish this through the inhibition of osteoclast formation from immature precursor cells and by simultaneously inducing apoptosis in mature osteoclast cells [14–16]. Bisphosphonates have also been found to have a direct effect on tumor cells; recent studies indicated that aminobisphosphonates induce apoptosis in a

72

variety of neoplastic cells, including myeloma and breast carcinoma cells. The mechanisms involved include interference with the mevalonate pathway [17–19], which in turn blocks protein prenylation and promotes the activation of caspases [18, 20]. This manifest antitumor activity and apoptosis induction ability [21] combined with an observed paucity of side effects in clinical trials make bisphosphonates a strong contender in the provision of adjuvant therapy against GCT [22]. In the present study, we investigated the potency of bisphosphonate-mediated apoptosis in both osteoclastlike giant cells and stromal tumor cells by using surgically obtained GCT specimens and GCT primary cultures.

Methods Reagents Alendronate (4-amino-1-hydroxybutylidene) was a gift from Merck Sharp and Dohme Ltd (Ireland). Pamidronate (3amino-1-hydroxy-propylidene-1, 1-bisphosphonate) and Zoledronate (1-hydroxy-2-[(1-himidazole-1-yl)ethylidene] 1-bisphosphonate) were obtained from Novartis Pharmaceuticals Ltd. (Basle, Switzerland). Stock solutions of bisphosphonates were prepared in phosphate-buffered saline, adjusted to pH 7.4, and sterilized by filtration. Other chemicals and reagents were of an analytical grade. GCT Specimens and Primary Culture The study was approved by the local ethics committee (Prince of Wales Hospital) and twelve patients were involved in pamidronate therapy and signed an informed consent form agreeing to the use of surgical specimens for research purposes. Six additional patients refused any form of adjuvant therapy but agreed to participate as controls in a clinical outcomes study. These parents were treated with surgical curettage only. The biopsy specimens was obtained from patients with GCT before commencing clinical treatment with pamidronate. Pamidronate was administered intravenously in a 90-mg per week dosage for 6 weeks. Afterwards the tumor was removed. GCT specimens from both the biopsy and resection were formalin fixed and paraffin embedded, and consequently sectioned. For the establishment of a GCT primary culture, GCT tissues from the biopsy or resection were freshly minced in Dulbecco’s minimum essential medium (DMEM) containing 100 U/mL penicillin and 100 lg/mL streptomycin. The resultant suspension, together with small tissue fragments, was then transferred to 25-cm2 flasks for subsequent culturing at 37C in 5% CO2 and 95% air overnight. Half the medium was changed on the second day. The culture was further maintained in DMEM supplemented with 2 mM L-glutamine, 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 l/mL, streptomycin until confluence. Primary cultures obtained after the ninenth passage, which represent the proliferating homogenous tumor cell population, were used for bisphosphonate treatment and evaluation. TUNEL Assay TUNEL assay (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) was used to assess apoptosis through the ApopTag Peroxidase In Situ Apoptosis Detection

S. M. Kumta et al.: BPs Induce Apoptosis in GCT

Kit (Intergen, Purchase, NY, USA). Briefly, paraffin-embedded sections were dewaxed in xylene and rehydrated through descending concentrations of A-grade alcohol and finally washed with distilled water. The sections were subsequently digested by 20 lg/mL proteinase K (Sigma chemical) for 15 minutes. The resultant was washed three times in phosphate buffered saline (PBS) and then treated with 3% hydrogen peroxide hydrogen peroxide (H2 O2) for 5 minutes to inhibit endogenous peroxidase activity. Treated sections were washed thrice more with PBS, soaked in an equilibrium buffer (provided from the kit) for 10 minutes and incubated at 37C for 60 minutes in a moist chamber with 30% TdT (terminal deoxynucleotidyl transferase) enzyme in a 70% reaction buffer. The reactions were stopped with stop/wash buffer for 10 minutes with shaking. The sections were rinsed three times with PBS for 1 minute at room temperature. This was followed by the addition of anti-digoxigenin peroxidase conjugate to enable marker visualization for light microscopy. Once the substrate 3,3¢-diaminobenzidine tetrahydrochloride (DAB) was added and sections were stained, the sections were washed, dehydrated, counterstained, and mounted. Their cellular morphology was assessed by light microscopy at ·200 magnification. For quantification of apoptosis, stained cells were counted from 20 random selected views by using the Metaphore image analysis program. The results were presented as an apoptotic index (number of apoptotic cells over total number of cells counted). In Vitro Drug Treatments and Flow Cytometry GCT primary cultures were seeded in 6-well plates at a density of 2.5 · 105 cells per well in DMEM/10% FBS overnight. Cells were then treated with alendronate, pamidronate, or zoledronate in different concentrations (0, 3, 30, and 150 lM) for 48 hours. Following bisphosphonate treatment, cells were harvested using 0.25% trypsin-EDTA (Gibco, Grand Island, NY, USA) and then washed 3 times in PBS. Floating cells that detached from culture vessels after bisphosphonate treatment were also collected by centrifuging the culture medium. Combined cell pools were finally resuspended in 250 lL of labeling solution according to the manufacturer’s instructions and incubated for 10–15 minutes in the dark. Labeled cells were then counted on a flow cytometer (FACScan, Becton Dickenson, Bedford, MA, USA) within 30 minutes. AnnexinV-FITC labeling was measured at 518 nm on the FL-1 channel (FITC-detector) and propidium iodide (PI) staining at 620 nm on the FL-2-channel (phycoerythrin-detector). Cell population was indicated in Figure 1A, R2 and R3 were used to calculate the total cell death. The mean percentage of total cell death from 12 primary cultures’ treatment with bisphosphonate were summarized and used in the calculation of statistical significance from an SPSS (spss 11.0 for Windows, one-way ANOVA (Chicago, IL, USA)).

Results The Induction of Apoptosis Following Bisphosphonate Treatment in Clinical Specimens

TUNEL labeling is indicative of genomic DNA fragmentation and is unevenly localized in each apoptotic nucleus. As shown with arrows in Figure 2, the nuclei in the multinucleated giant cells and stromal tumor cells showed various intensities of TUNEL, which represent different sensitivities to pamidronate. The numbers of apoptotic stromal cells and osteoclast-like giant cells were quantified by using an image analysis program. The apoptotic indexes of 12 cases are summarized in

S. M. Kumta et al.: BPs Induce Apoptosis in GCT

73

Fig. 1. Flow cytometric analysis of bisphosphonates induced apoptosis in GCT primary cultures. A, A representative flow cytometric analysis of pamidronate (P) induced apoptosis in one GCT cell line. R1 = Unstained cells; R2 = annexin-V binding cells (apoptotic cells); R3 = annexin-V and PI-stained cells (necrotic cells); R4 = PI stained cells. B, Summary of

flow cytometry results. Cells were treated with alendronate (A), pamidronate (P), and zolendronate (Z) at different dosages—0, 3, 30, or 150 lM for 48 hours. Experiments were run in duplicate. The mean percentages of apoptosis were plot from 12 primary cultures. (mean ± SEM, n = 12, P < 0.01 was considered statistically significant).

Table 1. As shown, in all cases, apoptotic indexes in both types of cells exhibited a marked increase after pamidronate treatment, ranging from 7.6% to approximately 54% in the stromal tumor cells and 29.7% to approximately 73.9% in the giant cells.

were more pronounced at higher concentrations of the drugs.

The Induction of Apoptosis in GCT Primary Cultures Morphologic Changes. Following incubation of GCT cell

lines with bisphosphonates for 48 hours, morphologic changes became prominent. As shown in Figure 3, treated cells changed their shape from spindle-like to a round or more oval shape (indicated with arrows), and began to detach from the culture vessels. These changes

Binding of Annexin-V to phosphatidylserin of cell plasma membrane characterizes an early and characteristic step in the apoptosis pathway. Annexin-V-Fluos and (PI) staining followed by flow cytometry analysis was essential in confirming and quantifying apoptosis after various bisphosphonate treatments. Each GCT primary culture exhibited an individual response to the drugs. To summarize, the efficacy of alendronate, pamidronate, and zoledronate in inducing total cell death is shown in Figure 1B. The mean percentage of total cell death in the control cultures was 4.68%. Zoledronate was the

Quantitative Analysis Using Flow Cytometry.

74

S. M. Kumta et al.: BPs Induce Apoptosis in GCT

Fig. 2. Photomicrographic images illustrating pamidronate-induced apoptosis in GCT by use of TUNEL assay. TUNEL-labeled DNA fragmentation was indicated with arrows in sections prior to pamidronate treatment (A) and postpamidronate treatment (B). Typical TUNEL-labeled DNA fragmentation was shown in higher magnification (B, insert). Original magnification ·400.

Table 1. Pamidronate-induced apoptosis assessed by TUNEL assay Stromal tumor cell

Osteoclast-like giant cell

Case

Before pamidronate

Pamidronate

Before pamidronate

Pamidronate

G1 G2 G3 G4 G5 G6 G8 G10 G12 G13 C7 C9

17/1320 12/1246 11/1163 21/1492 — 39/1253 42/1344 13/1475 14/1432 — 11/1325 36/1534

549/1654 374/1316 232/1735 131/1720 603/1658 232/1383 211/1612 735/1354 327/1536 521/1598 440/1637 572/1754

5/687 284/854 4/647 3/571 — 16/648 65/754 4/874 27/746 — 45/687 160/754

286/681 631/854 352/762 403/745 216/638 479/867 464/785 207/698 332/749 425/812 364/695 539/845

(1.3%) (